THE ORGANIZATION AND MANAGEMENT OF

THE NORWEGIAN SOCIAL SCIENCE DATA SERVICES (NSD)

JARLE BROSVEET, BJ^RN HENRICHSEN^ LARS W. HOLM, AND TERJE SANDE
NORWEGIAN SOCIAL SCIENCE DATA SERVICES

BERGEN, NORWAY

(This paper was deliveved at the 1981 IFDO/IASSIST Conference, Grenoble.)

THE EARLY YEARS

The origins of the Norwegian Social Science Data Services (NSD) go

back to 1967 when the Norwegian Research Council could no longer ignore

various complaints about, long delays and frequent errors in the processing

of social science data. ' As a result the Council decided to set up a

committee to study the reasons why bottlenecks occurred in computing

services at the universities. The committee was also instructed to make

proposals for improvement of social science computer facilities and

services.

The committee made a number of recommendations and took the first steps

to implement these as part of its work. A series of intensive courses in

computing for social scientists was organized, a variety of statistical

packages and specialized programs were acquired and installed on the

university computers, and steps were taken to build up a first set of data

banks easily accessible to researchers at all universities.

In 1970, the committee proposed to institutionalize these activities

by establishing a nation-wide organization supplying various types of

services to all social scientists. The Research Council approved the

proposals and established the NSD for a trial period through 1974. Later

a new program of activities was approved for the next three-year period.

As the activities proved a success, the NSD was given permanent status

from January 1st, 1978.

1)
For background information, see Stein Rokkan, "National Primary

Socio-Economic Data Structures III: Norway", International Social

Science Journal 30(3), 1978:621-652; also Stein Kuhnle and Stein

Rokkan, "Political Research in Norway 1960-1975", Scandinavian
Political Studies 12, 1978:127-156.

73



THE ORGANIZATION

The NSD differs from most other organizations of this type in three
ways

:

- it is a federally structured organization with offices at four
universities and with close working arrangements with the regional
colleges;

- it has built up a wide variety of data resources across all fields
of the social sciences: the data holdings comprise surveys as

well as large databanks for communes and census tracts, an archive
of information about organizations, and a series of computerized
files of data on the recruitment and careers of various elite groups;

- it has established close contacts with a wide range of users in the
social sciences and in administrative bodies, partly through its

regular newsletter (Brukermelding) , partly through the annual meeting
of representatives of a great majority of research institutes and
teaching departments active in the social sciences throughout Norway.

The NSD has its headquarters at the University of Bergen. In addition
there are external data secretariats at each of the other universities:
Oslo, Trondheim and Troms0.

THE DATA

The NSD has so far built up seven basic types of data holdings:

I The first and largest is the database for communes . This database
covers statistics for all local units of administration since 1800 and
is linked up with a facility for computer cartography (POLYVRT,
CALFORM, SYMAP, FIGUR). This is our most widely used facility and is

constantly expanded and improved. A great amount of work has been
expended on finding effective solutions to the problems posed by
changes in boundaries and units . The database offers detailed
documentation of these changes and provides the user with a variety
of coefficients for recalculating data values whenever changes occur.

II The NSD has set up a cartographic service useful to a number of users.
Coordinate matrices for all commune boundaries in Norway since 1800
have been computerized. The boundary segments are identified by a

time key to allow the production of maps for every year since 1800,
Initially, maps were mostly conformant using cross-hatchings to
represent the data values. Later programs have been developed to
display information at the centerpoint of every commune. This
development makes it possible to display not only multivariate
statistics, but also discrete data about the presence or absence of
particular infrastructure elements such as airports, harbours,
hospitals, schools, factories etc.

III To allow analyses at a lower level of aggregation the NSD has taken
steps to organize a set of data at the lowest level of official

74



enumeration: the census tract . This data bank contains data from the
censuses of 1960 and 1970 and has proved of great interest to city
planners, geographers, anthropologists and sociologists.

IV In dealing with survey data it has been our policy to link up response
data for comparable questions over time , so we have deliberately de-
clined to serve as a depository for individual studies. The largest
file so far developed covers the surveys of the Norwegian Gallup
Institute (roughly monthly ones) from 1964 through 1976. Currently,
this archive is being updated with data from the Morsk Opinionsinstitutt
A.S. (Norwegian Opinion Institute). A similar linkage is being
completed for the electoral surveys carried out by Professor Henry
Valen and his colleagues since 1957. Some of the most thorough surveys
carried out by the Central Bureau of Statistics since 1967 are now at

the disposal of academic users under an agreement with the NSD. The
NSD has prepared detailed documentation for some of these surveys and
reformatted them for use with SPSS to increase their availability.

V A major effort has been made to develop a systematic databank for elite
groups. The first file to be completed contains biographic information
on Members of Parliament since 1814 , This archive gives information
on father's occupation, education, early career, positions in legis-
lative conmittees etc. The file has recently been linked with one for
roll calls since the establishment of party fronts in the 1870's to

allow analyses of interrelations between background factors and
legislative behaviour. Similar biographical files are being organized
for members of the central administration as well as for graduates
from the universities since 1811.

VI Another file taken over by the NSD is the archive of information about
voluntary associations in Norway 1971/1972 .

VII The NSD has also got a file of test data for the recruits to the Armed
Forces from 1951 to 1968.

In organizing these as well as other data the NSD has emphasized the

paramount importance of documentation and data linkage . The NSD does not
wish to serve as a pure depository of data. Consequently, data files are
rarely incorporated into our holdings unless they can be fully documented
and linked up into a system either topically or across time periods.

To make sure that our facilities are used extensively, the NSD has

given high priority to educational activities , courses, lectures and
teaching packages. Active support has been offered within the programme
of the International Social Science Council as one of the cross-national
workbooks in the ISSC series has been produced in close cooperation with
the NSD. Steps have also been taken to compile a series of specifically
Norwegian teaching packages. One of these is based on fiscal -administrative
data for communes , another focuses on time-series data for intermediary
levels of regional aggregation (fylker ). Further packages are based on

data from electoral surveys and the file of biographical information on

Members of Parliament.

75



MATTERS OF POLICY AND ECONOMY

Up until 1977 NSD was almost exclusively financed by grants from the

Social Science division of the Research Council. At that time it was

realized that NSD holdings were used extensively by groups other than social

scientists and for purposes other than research and teaching. Even more
important, some governmental agencies as well as the NSD realized that

our data archives carried a huge potential for several kinds of governmental
planning and research. This potential was likely to grow not least as a

result of increased co-operation with the Central Bureau of Statistics.
Consequently, our financing had to be altered in one way or another.

Several models were discussed, but in the end the NSD is still being financed
on a project basis. A major advantage of this arrangement is that it

allows the Research Council and the NSD to set the priorities. At the same
time we can also choose the governmental projects that appear to give the
greatest benefits to the social sciences in general.

From 1978 to 1981 the basic grant to the NSD increased from about 1

mill, to 1.4 mill. Nkr. , each year totalling about 8% of the Research
Council allocation to social science projects. If this were to be our
total budget, it would restrict our expansion considerably. However, we
have been able to obtain supplementary funding elsewhere, including grants
from other divisions of the Research Council, the Nordic research councils
and various governmental agencies.

The current funding of tne NSD allows us to keep a reasonably large
group of professionals permanently employed. As of today all full-time
staff except the secretaries have a university degree. Our recruitment
policy has always been to emphasize social science background with experience
in data analysis and programming. Such experience is achieved through the
training of recruits as assistants either at the NSD or on large research
projects at the universities. The academic background of the staff covers
almost the full range of the social sciences: economic:, sociclccy,
political science, geography, public administration, history and infor-
mation science. Not only does this provide a good interdisciplinary
working group, it also makes it easy to communicate with various research
interests. The recruitment policy definitely reflects the tasks that have
been assigned priority by the Board of the NSD.

To put it simply, data and documentation are given priority over
software development and programming. However, the NSD is seldom involved
in primary data collection. In effect, we place great store by
establishing good working relations with data collecting agencies in

administrative bodies, private firms and research institutes.

Obviously^ the most important of these is the Central Bureau of
Statistics. We now cooperate rather extensively with the Bureau and also
exchange services to some extent. This co-operation means that we are
able to provide social scientists with virtually all data collected by
the Bureau at a very low cost that often amounts to no more than the
copying of the magnetic tape. The NSD has also been assigned the respon-
sibility for giving social scientists access to Bureau data that are not
completely anonymized, provided that the Bureau receives feedback as to
who is given access. Furthermore, the NSD is represented on committees
to review census questionnaires and Bureau tables to be publicized using
census iata.

76



To a certain extent we are developing similar relationships with

other governmental agencies supporting data collection tasks.

Currently, we are cooperating with the Ministry of Consumer Affairs

on the computerization of a register of representatives in official boards

and councils. We do part of the coding and punching as well as the

running of simple tables to be used in the official report. Later, these

data will be established as an archive and included in the data holdings

available to researchers free of charge.

There are several reasons why such excellent and confident relation-
ships can be developed vis-a-vis the governmental agencies. One of the

most important reasons seems to be that there has never been any abuse of

data by students or researchers. The NSD has also shown the ability to

store and document data in a way that permits simple computer runs, map

drawings etc. much faster, cheaper and often with better quality compared

with other data distributors. Since the holdings of the NSD contain data

collected by means of public funds, there is also a strong argument for

maximum public usage of these data.

Before leaving our relationship with governmental agencies, we must

also mention the effect of privacy legislation on social research. As soon

as the privacy law was proposed, the NSD took the initiative to help

fulfill the intentions of the law as well as to counteract some of the

negative effects experienced in other countries. After lengthy negotiations
we are now in the process of implementing a strategy that we believe is

advantageous for all parties concerned. First, the NSD has been given a

general consession by the Data Inspectorate to archive and store relevant

social science data. Second, we have obtained a grant for acting as a li-

aison between the research community and the Data Inspectorate. It also

enables us to set up a secretariat for the Research Council on matters

concerning privacy and data collection.

INTERNATIONAL COOPERATION

On the international side, the NSD has always been cooperating
actively with similar organizations in other countries. It was one of the
founder members of the International Federation of Data Organizations in
1977 and is active in a Working Group under its auspices for the
coordination of local - regional data bases and computer cartography .

Since 1971, information about European archiving efforts has been
published regularly in the European Political Data Newsletter set up by
Stein Rokkan under the European Consortium for Political Research and
later co-sponsored by the Norwegian Social Science Data Services. The
EPD Newsletter has developed a classification scheme for the comparison
of the contents of European time-series datasets and has repeatedly
advocated actions to co-ordinate developments in this field.

First, NSD and the Norwegian Research Council succeeded in securing
the sponsorship of the European Science Foundation for a meeting on
"Databases for Regional Analysis" in 1977.

77



Second, NSD took steps to develop a proposal for the establishment of

a Joint Nordic Database for Regional Time Series during 1977-1979.

Funding for a three-year period was granted by the four Nordic Social

Science Research Councils, i.e. the Danish, Finnish, Norwegian and

Swedish. The project started in January 1979.

Also, at the IFDO meeting on regional data held in Turin in March

1980, the task of coordinating an effort to summarize and structure

information for a joint European database was entrusted to NSD. NSD was

asked to complete the matrix or map of available data on the basis of the

work started in Turin, to disseminate a report and organize a workshop in

which further action can be discussed.

The effort implies the following:

a. to identify organizations and scholars in each nation who are

willing to cooperate, and to look for additional researchers and

scholars in cases where regional or thematic gaps in the databases

have been identified;

b. to establish a cornnon standard for types of variables in these

databases;

c. to establish a common standard for documentation of these data;

d. to carry out an effort as far as possible to establish the

comparability of these data. This task does not necessarily imply

the establishment of formal statistical comparability, but a more

substantive kind of theoretical comparability, such as the

possibility of using analogous measures for common research
projects of analysis of social structures and processes.

NSD concludes from this information-gathering effort that it will be

feasible to start the building up of a first version of a Joint European
Data Base for Regional Time Series. At this time we have established a

network of contacts and compiled information about data matrices and

variable lists. A report on the progress of the project has been submitted

to the Board of IFDO for further discussion.

78